Hacker News Reader

Comment by t0mas88 | Hacker News Reader

We had a Glassdoor sales guy contact HR after a negative review to sell us a paid option to "improve" our profile. My trust in them since then is below zero.

View on HN

To be clear, I don’t really trust Glassdoor that much either, but I think more details are warranted here. If you can alter the reviews by hiding them or changing the ranking algorithm, that’s basically fraud. On the other hand, if only paid company accounts can get access to additional analytics / reporting, have the ability to respond inline to reviews, add additional info to the company profile, etc, then that’s less problematic.

So which one Glassdoor is then? Do they allow companies to hide/improve reviews for premium pay?

Or is it just additional features of the platform without altering the review data?

Ah the Yelp business model.

I, like you, used to flippantly accuse Yelp of extortion. It’s always easy to believe the worst. Then someone here on HN provided several helpful articles that gave me pause and helped me update my point of view.

Yelp themselves have addressed this with a dedicated page: https://www.yelp.com/extortion

Maybe I’ve been duped. I’d be curious if your comment is based on different research, or personal experience.

That page would be much more convincing, if it listed out what had Yelp done for local businesses in exchange for money. In modern corporate PR, most of the useful information is in what is not said.

I have a few friends who own restaurants. Yelp extortion is real. i've seen it myself. Yelp denying it is obvious, who'd admit to it?

They do not "manipulate" ratings, sure... They just take a looong time to "Review" positive reviews, and not too long to "Review" negative reviews you get. And, of course, only reviews they have "reviewed can be shown and count towards your average score...

All these claims are publicly testable using web archives to build stats about review approval speed. Did you verify these claims?

>All these claims are publicly testable using web archives

I'm confused about how you think this would work.

1. Yelp doesn't report on its approval speed.

2. To get an accurate report on this, you would have to have several accounts posting positive and negative reviews on the same pages. Yelp doesn't allow this and would ban the accounts.

3. Yelp explicitly disallows scraping their site (i.e. having a bot archive the pages). Doing so would be thwarted by their anti-bot measures. It's also against their TOS.[0]

Do you have web archives that can be used to build these stats?

[0]: https://www.yelp-support.com/article/Can-I-copy-or-scrape-da...

For starters, i don’t even have to give you any hints because if you’re claiming somebody’s doing extortion you better have some good evidence.

1. You don’t need that to demonstrate statistically significant difference between paying and non-paying customers

2. You don’t need to make any accounts, only read existing and incoming data

3. Yes, you shouldn’t break TOS but you also shouldn’t be throwing around unsubstantiated accusations of extortion

I have seen reviews left by customers with my own eyes. Literally watched them type them and send them. I then watched how long they took to appear.

I have no need to convince you or anyone else. But I am convinced. I trust my eyes a lot more than I trust yelp.

And did you compare that to how long it takes to approve reviews to other businesses? Paying and nonpaying?

Right now you don’t even have proper N=1.

There are endlesss stories of yelp extortion from so so mamy independent sources. That yelp needs a top level page to address this is proof enough. There is zero amount of corporate speak from Yelp that would change my mind.

What’s your evidence?

Extortion as a service.

Wow, I didn't realize they operated like yelp. Are they hiding reviews or removing them outright?

A former company I worked at easily and quickly was able to have Glassdoor remove a post from a disgruntled former employee that revealed some private information about the company (sales numbers or something).

Good old extortion

Reply on news.ycombinator.com